Text this: Learning language to symbol and language to vision mapping for visual grounding

           __   __   ______     ______   __   _   
    ___    \ \\/ // |      \\  /_   _// | || | || 
   /   ||   \ ` //  |  --  //   -| ||-  | '--' || 
  | [] ||    | ||   |  --  \\   _| ||_  | .--. || 
   \__ ||    |_||   |______//  /_____// |_|| |_|| 
    -|_||    `-`'   `------`   `-----`  `-`  `-`  
     `-`